The only real risk is that some day some specification or some browser or some web-wide indexing robot might start treating noindex as a real element with some defined meaning, possibly with default rendering and default functionality. Then you would be in trouble if these differ from what you expected. But its a rather small risk, and it seems that you dont have a choice.
There is nothing you need to do. Browsers are expected to ignore unknown tags, and they do, so they see <noindex>foo</noindex> just as foo. Well, not quite. Technically, modern browsers construct an element node (of type HTMLUnknownElement) in the DOM. But the element has no associated default styling and no associated action, so its really a dummy element and represents its content only.